We meet South Korean serial entrepreneur Chester Jungseok Roh to ask how to do it, and the sacrifices it takes.
Roh has founded six companies, mainly within software. Currently, he is the CEO of BFACTORY, the Seoul-based company behind skincare brand KYYB and make-up brand AMELI, and is leveraging AI technology across the business. At Pitti Fragranze in Florence, he took the stage to share the current state of new tech in beauty and fragrance, and we caught him afterwards.
In general, do you think beauty brands are good enough to use AI and new technologies?
– Not at all, he states.
Why not?
– The industry is far from what’s happening in the technology scene. Beauty and fashion are normally areas that are adapting technology at the very latest stage. However, I think we are entering a wholly new era in everything that is combined with AI. What’s happening in the technology world will happen in beauty and fashion pretty soon. That’s my expectation.
Primarily, in beauty, the use of AI has been more in R&D, finding new ingredients, and exploring new formulations. But you mean to take a more holistic view of it?
– Sure. We can apply AI to everything in every industry. Not only in the making of research and what you would call the really, really heavy stuff. All the light stuff that has been handled by humans can also be done by AI. When I’m speaking of how AI can help create an autonomous company, the definition of such a company is not that I won’t be hiring any humans at all. The actual meaning is to stop the everyday, mundane, useless things and move into working with the more creative and value-creating areas as a human. That is what I mean.
Can you take a very concrete example, the one thing that first pops up in your mind, of how AI saves time and resources in your daily operations?
– The job of the knowledge worker is just one thing, or maybe the combination of several tasks. But most of the main task is actually bringing the data in, analysing it, and handling Excel, PowerPoint, and the documentation. It’s pretty laborious. Actually, most of the job currently for humans, up to 80%, is these types of tedious, mundane, laborious things. We can free people from that burden and move them into a better one.
– I know that there are people who like this concept, while most of them may not like it. Look at what Silicon Valley is already doing with ChatGPT and Claude: in software development, AI agents now take on whole tasks—they write the code, test it, and report back—and the engineer’s job has shifted to giving direction and checking the result. That is not a distant future; it is already happening there. And what happens in software will happen in every industry, beauty included, so we should prepare ourselves now.
Your ambition is to create an autonomous beauty company. Why?
– ‘Why’ is a good question. Other than spending time on less essential things, we wish to relieve and eliminate every other aspect of the conventional job, although that might be a bit of an evil thing to say as an entrepreneur.
How far have you come in this journey to become autonomous?
– I think that I can say around 70% of our overall operations.
What’s left?
– Combining the data pipeline in more detail. Today, humans are doing the decision-making, and someday, maybe, AI can handle that on our behalf, while the remaining job of a human being is to just go out and play with other people. That day will come, I guess. That’s why many of the erudite people in Silicon Valley are talking about universal basic income. The job of people is just to go out and play and create more love in the world. That might be an answer to that question.
When would you expect this to happen?
– Within ten years, 100%. Within five years, maybe a 30 or 40% possibility.
Apparently, foreign beauty CEOs come and visit your office. What surprises them the most?
– The biggest one is that they actually don’t know that AI is good enough to handle so many jobs. That’s the one thing they are most frightened of and amazed at.
And you’re also gonna offer them special B2B software services to support them in this process.
– Yes, it’s called NOVA. We already run it in-house at BFACTORY, and the external B2B launch is planned for Q4. We are working closely with Cosmax, the world’s largest cosmetics ODM, where I also serve as AI Adviser. They have a lot of customers already. Together with them, we plan to offer NOVA to its brand customers alongside Cosmax’s existing services.
From an outside perspective, K-Beauty is a world leader in beauty innovation. Are you also saying that South Korea can be a leader in new tech, where China is so strong?
– Frankly speaking, I just mentioned on stage the importance of the speed of change in K-Beauty. From that perspective, we do have competitiveness.
– But, as you just mentioned, China is following this. In the speed of product development, I think they are now almost at the same level as Korea. In building trusted, premium global brands, Korea is still ahead, so the speed of change alone would not solve the problem.
– As a brand, we should move our position to a more premium one rather than the mass and affordable price range. How shall we make the premium K-Beauty brand that might sustain the next 30 years and become L’Oreal, Chanel, or any other famous French brand? That would be the grand challenge of K-Beauty in the next several years.
That might be your seventh enterprise to start.
– It’s actually what we’re trying to make with KYYB. We position ourselves as a K-Beauty premium brand in the States first, and then go to China. Europe is a market we are actively exploring, and I’m also expecting some market share there.
On stage, you also mentioned that we’re approaching the end of distribution. What do you mean?
– For conventional distribution, there are a lot of other channels, and each channel has its own characteristics. By managing the channel strategy properly, you could get some benefits. But the distribution channel is becoming monolithic. What I mean by that is that ChatGPT, Gemini, and AI agents would be the sole endpoint for customers to get information and to buy the product. I think that is the huge problem, and that’s what I call the end of distribution. This monopoly or oligopoly situation of distribution would happen because of AI.
– Agent shopping, this ‘end of distribution,’ is not a prediction for us; it is already happening. In our make-up brand AMELI, customers can shop through HEVA, our AI personal shopper, which remembers what they have bought and asked before and tries looks on their own face. Today, 7% of all our purchasing customers buy through HEVA’s in-depth AI recommendations, and their retention is far higher than that of our other customers. For brands, that is the real lesson: if customers are going to shop through AI agents anyway, the brands that build the best agent experience themselves will win those customers back directly.
This will change quite a lot.
– Yes. The power of the incumbent distributors will shift to whoever owns the AI interface, says Jungseok Roh. We are entering wholly new territory, and yes, a lot of new things will happen!
Words: Johan Magnusson